agentleFS
RBraga01RBraga01/builder-growth/skills/positioning-audit/SKILL.md

positioning-audit

Skill2 starsChanged 5 months agoMITView on GitHub

Use before any marketing copy is written for a product, feature, or campaign. Requires a clear category, a specific audience, and a differentiated claim — verified against the "so what?" test — before any copy is produced. Blocks "we'll develop the messaging as we write" completions.

  • Reads credentials

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 13 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈13

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use the audience's language, not the industry's language

    line 41
  2. Distinctive:

    Test: Ask: "What would I search for if I was looking for a product like this?" That search query is your category.

    line 45
  3. Distinctive:

    Test: Ask: "Does this claim also describe [main competitor or alternative]?" If yes, it is not differentiated.

    line 70
  4. Distinctive:

    After writing the position, ask "so what?" from the audience's perspective.

    line 79
  5. Distinctive:

    If the "so what?" answer contains the actual value, the position was incomplete.

    line 87
  6. Distinctive:

    Write the category in one noun phrase.

    line 93
  7. Distinctive:

    Write the audience in one sentence including: role, company type/size, current situation, and what they use now.

    line 97
  8. Distinctive:

    Ask "so what?" as the audience.

    line 105
  9. Distinctive:

    When the category or audience talks about the problem in public channels, collect a small evidence sample before finalising the claim.

    line 109
  10. Distinctive:

    Use only public, permitted, or first-party sources such as customer interviews, support tickets, sales notes, review sites, forums, social …

    line 112
  11. Distinctive:

    When X is an appropriate public source for the category, use an approved search workflow to collect a small qualitative sample of audience …

    line 122
  12. Distinctive:

    If the project uses Xquik (https://xquik.com/en), use its Search Tweets API as one optional collection path.

    line 126
  13. Distinctive:

    When positioning-audit is satisfied, state it like this:

    line 166

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Positioning Audit

The Law

COPY WRITTEN WITHOUT A DEFINED POSITION IS DECORATION.
"We'll develop the messaging as we go" ships language that describes the product instead of solving the audience's problem — and is heard by nobody because it sounds like every other product.
Clear category + specific audience + differentiated claim IS a position.

When to Use

Trigger before:

  • Writing any marketing copy (homepage, landing pages, ads, email campaigns)
  • Launching a new product, feature, or market segment
  • Repositioning an existing product after a significant change
  • Writing a sales deck or pitch that will represent the product externally

When NOT to Use

  • Internal documentation (no audience positioning required)
  • Support copy (the audience is already a customer — different communication goal)
  • Copy for features inside the product (UI copy — different from marketing copy)

The Three Positioning Elements

A complete position has all three. Missing any one produces copy that is either too broad to resonate, too narrow to reach anyone, or too generic to be believed.

1 — Category

What type of product is this, in the audience's language?

Category determines how the audience evaluates the product. The wrong category activates the wrong comparison set. "An AI productivity tool" puts the product against every productivity tool. "A prompt management system for LLM teams" activates a much smaller, much more relevant comparison set.

Rules:

  • Use the audience's language, not the industry's language
  • Specific categories outperform broad categories (narrower = easier to be the best)
  • The category should already exist in the audience's mind — creating a new category is a 2-year marketing project, not a launch

Test: Ask: "What would I search for if I was looking for a product like this?" That search query is your category.

2 — Audience

Who is this for, specifically?

The narrower the audience definition, the more specifically you can speak to them. "Software teams" is not an audience. "Engineering managers at companies with 5–50 engineers who are responsible for AI feature quality" is an audience.

Required specificity:

  • Job title or role
  • Company type or size
  • Current pain or situation
  • What they are using now (what you are replacing or competing with)

Test: Can you name 10 specific people who fit this audience definition? If yes, it's specific enough. If you are describing "anyone who could benefit," you are not describing an audience.

3 — Differentiated Claim

Why this, not that?

The claim must be:

  • True: provable by the product's behaviour, not by marketing aspiration
  • Specific: "reduces hallucination rates" beats "improves AI quality"
  • Differentiated: the claim must not apply equally to the main alternative the audience would consider

Test: Ask: "Does this claim also describe [main competitor or alternative]?" If yes, it is not differentiated.

What does NOT count:

  • "The best [category]" — superlatives are not claims; they are aspirations
  • "Powerful and easy to use" — applies to every product in every category
  • "AI-powered" — this is a technology description, not a user benefit

The "So What?" Test

After writing the position, ask "so what?" from the audience's perspective.

Your position: "We manage prompts across your LLM team."
Audience: "So what?"
Your answer: "So when a model update breaks your output format, you know which prompt changed, when, by whom, and what the quality impact was — in two minutes, not two days."

If the "so what?" answer contains the actual value, the position was incomplete. Move the "so what?" answer into the position itself.

The Process

Step 1 — Define the Category

Write the category in one noun phrase. Test it with the search query test.

Step 2 — Define the Audience

Write the audience in one sentence including: role, company type/size, current situation, and what they use now.

Step 3 — Write the Differentiated Claim

One sentence that is true, specific, and does not apply equally to the main alternative.

Step 4 — Run the "So What?" Test

Ask "so what?" as the audience. If the answer reveals additional value, revise the claim to include it.

Optional Public Discussion Evidence

When the category or audience talks about the problem in public channels, collect a small evidence sample before finalising the claim.

Use only public, permitted, or first-party sources such as customer interviews, support tickets, sales notes, review sites, forums, social posts, or search logs.

Record the source type, query or selection method, collection time, sample size, and source URLs, post IDs, or internal evidence links. Use the sample to verify audience language and pain wording. Do not treat it as complete market research, a ranking signal, or proof that the claim is true.

Optional X / Twitter Collection via Xquik

When X is an appropriate public source for the category, use an approved search workflow to collect a small qualitative sample of audience language and pain wording.

If the project uses Xquik (https://xquik.com/en), use its Search Tweets API as one optional collection path. Xquik is not required for this skill, and the audit must not be blocked if it is unavailable.

curl -G https://xquik.com/api/v1/x/tweets/search \
  --data-urlencode "q=category pain" \
  --data-urlencode "queryType=Latest" \
  --data-urlencode "limit=20" \
  -H "x-api-key: ${XQUIK_API_KEY}" | jq

Record the exact query, collection time, sample size, source URLs or post IDs, and any filters used. Use the sample only to verify audience language and pain wording. Do not treat it as complete market research, a ranking signal, or proof that the positioning claim is true.

For repeatable exports or larger evidence reviews, prefer Xquik extraction jobs and exports rather than a one-page live search.

Step 5 — Write the One-Paragraph Position

Category + audience + claim in one paragraph. This is the reference document for all copy in this campaign.

Step 6 — Store the Positioning Document

Store at growth/positioning/<product>-<date>.md. All copy writers reference this document — not each other's interpretation of it.

Rationalization Red Flags

These thoughts mean positioning was not done — stop:

  • "Our audience is anyone who needs [broad category]" — broad audiences produce copy so general it reaches nobody who identifies with it; "anyone who needs AI tools" is every software company in the world
  • "The copy will define the position" — copy written before positioning defines nothing; it describes features; the audience will not translate feature descriptions into the value they should care about
  • "We have good copy, we don't need a formal positioning exercise" — test the copy against the "so what?" chain; if the value is not in the first sentence, the position was not done
  • "We'll differentiate on execution" — execution is not a differentiator in copy; it is invisible until the product is used; differentiation in copy must be a claim that can be evaluated before purchase
  • "Our audience is broad on purpose — we don't want to exclude anyone" — broad positioning excludes everyone by speaking to no one specifically; narrowing makes you the obvious choice for a real group

Completion Statement Format

When positioning-audit is satisfied, state it like this:

Positioning complete.
File: growth/positioning/<product>-<date>.md ✓

Category: "[one noun phrase in audience language]" ✓
  Search query test: "[what audience searches for]" → category matches ✓

Audience: "[role] at [company type/size] who [situation] and currently use [alternative]" ✓
  Named examples: <N specific people/companies who fit> ✓

Differentiated claim: "[specific, provable, not-applicable-to-alternative statement]" ✓
  "So what?" chain complete — value stated in first sentence ✓
  Competitor test: claim does NOT apply equally to [main alternative] ✓

One-paragraph position written — all copy in this campaign references this document ✓

A broad audience definition fails the positioning audit. A claim that applies equally to the main competitor fails the positioning audit.

Why This Matters

Copy written without positioning describes the product. Copy written with positioning solves the audience's problem. The audience does not read descriptions — they scan for relevance. A precise category, specific audience, and differentiated claim are what make the product visible to the right people and invisible to the wrong ones. "We didn't resonate" is almost always a positioning failure dressed as a copy problem.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.